Bayesian Light Source Separator (BLISS)
========================================
[](https://prob-ml.github.io/bliss/)
[](https://github.com/prob-ml/bliss/actions/workflows/tests.yml)
[](http://codecov.io/github/prob-ml/bliss?branch=master)
[](https://pypi.org/project/bliss-toolkit)
# Introduction
BLISS is a Bayesian method for deblending and cataloging light sources. BLISS provides
- __Accurate estimation__ of parameters in blended field.
- __Calibrated uncertainties__ through fitting an approximate Bayesian posterior.
- __Scalability__ of Bayesian inference to entire astronomical surveys.
BLISS uses state-of-the-art variational inference techniques including
- __Amortized inference__, in which a neural network maps telescope images to an approximate Bayesian posterior on parameters of interest.
- __Variational auto-encoders__ (VAEs) to fit a flexible model for galaxy morphology and deblend galaxies.
- __Wake-sleep algorithm__ to jointly fit the approximate posterior and model parameters such as the PSF and the galaxy VAE.
# Installation
BLISS is pip installable with the following command:
```bash
pip install bliss-toolkit
```
and the required dependencies are listed in the ``[tool.poetry.dependencies]`` block of the ``pyproject.toml`` file.
# Installation (Developers)
1. To use and install `bliss` you first need to install [poetry](https://python-poetry.org/docs/).
2. Then, install the [fftw](http://www.fftw.org) library (which is used by `galsim`). With Ubuntu you can install it by running
```bash
sudo apt-get install libfftw3-dev
```
3. Install git-lfs if you haven't already installed it for another project:
```bash
git-lfs install
```
4. Now download the bliss repo and fetch some pre-trained models and test data from git-lfs:
```bash
git clone git@github.com:prob-ml/bliss.git
```
5. To create a poetry environment with the `bliss` dependencies satisified, run
```bash
cd bliss
poetry install
poetry shell
```
6. Verify that bliss is installed correctly by running the tests both on your CPU (default) and on your GPU:
```bash
pytest
pytest --gpu
```
7. Finally, if you are planning to contribute code to this repository, consider installing our pre-commit hooks so that your code commits will be checked locally for compliance with our coding conventions:
```bash
pre-commit install
```
# Latest updates
## Galaxies
- BLISS now includes a galaxy model based on a VAE that was trained on Galsim galaxies.
- BLISS now includes an algorithm for detecting, measuring, and deblending galaxies.
## Stars
- BLISS already includes the StarNet functionality from its predecessor repo: [DeblendingStarFields](https://github.com/Runjing-Liu120/DeblendingStarfields).
# References
Mallory Wang, Ismael Mendoza, Cheng Wang, Camille Avestruz, and Jeffrey Regier. *Statistical Inference for Coadded Astronomical Images.* Machine Learning and the Physical Sciences workshop, NeurIPS 2022. [arXiv:2211.09300](https://arxiv.org/abs/2211.09300)
Derek Hansen, Ismael Mendoza, Runjing Liu, Ziteng Pang, Zhe Zhao, Camille Avestruz, and Jeffrey Regier. *Scalable Bayesian Inference for Detection and Deblending in Astronomical Images*. ICML Workshop on Machine Learning for Astrophysics, 2022. [arXiv:2207.05642](https://arxiv.org/abs/2207.05642)
Runjing Liu, Jon D. McAuliffe, Jeffrey Regier, and The LSST Dark Energy Science Collaboration. *Variational Inference for Deblending Crowded Starfields*, 2021. [arXiv:2102.02409](https://arxiv.org/abs/2102.02409)
Raw data
{
"_id": null,
"home_page": "https://github.com/prob-ml/bliss",
"name": "bliss-toolkit",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.10,<4.0",
"maintainer_email": "",
"keywords": "cosmology,blending,weak lensing,bayesian,ml,pytorch",
"author": "Ismael Mendoza",
"author_email": "imendoza@umich.edu",
"download_url": "https://files.pythonhosted.org/packages/17/42/d32d35deff38cb04d8526106f444346a3e4ffe74dbbe553313c4ab24a165/bliss_toolkit-0.3.tar.gz",
"platform": null,
"description": "\n\n\nBayesian Light Source Separator (BLISS)\n========================================\n[](https://prob-ml.github.io/bliss/)\n[](https://github.com/prob-ml/bliss/actions/workflows/tests.yml)\n[](http://codecov.io/github/prob-ml/bliss?branch=master)\n[](https://pypi.org/project/bliss-toolkit)\n\n# Introduction\n\nBLISS is a Bayesian method for deblending and cataloging light sources. BLISS provides\n - __Accurate estimation__ of parameters in blended field.\n - __Calibrated uncertainties__ through fitting an approximate Bayesian posterior.\n - __Scalability__ of Bayesian inference to entire astronomical surveys.\n\nBLISS uses state-of-the-art variational inference techniques including\n - __Amortized inference__, in which a neural network maps telescope images to an approximate Bayesian posterior on parameters of interest.\n - __Variational auto-encoders__ (VAEs) to fit a flexible model for galaxy morphology and deblend galaxies.\n - __Wake-sleep algorithm__ to jointly fit the approximate posterior and model parameters such as the PSF and the galaxy VAE.\n\n# Installation\n\nBLISS is pip installable with the following command:\n\n```bash\npip install bliss-toolkit\n```\n\nand the required dependencies are listed in the ``[tool.poetry.dependencies]`` block of the ``pyproject.toml`` file.\n\n# Installation (Developers)\n\n1. To use and install `bliss` you first need to install [poetry](https://python-poetry.org/docs/).\n\n2. Then, install the [fftw](http://www.fftw.org) library (which is used by `galsim`). With Ubuntu you can install it by running\n\n```bash\nsudo apt-get install libfftw3-dev\n```\n\n3. Install git-lfs if you haven't already installed it for another project:\n\n```bash\ngit-lfs install\n```\n\n4. Now download the bliss repo and fetch some pre-trained models and test data from git-lfs:\n\n```bash\ngit clone git@github.com:prob-ml/bliss.git\n```\n\n5. To create a poetry environment with the `bliss` dependencies satisified, run\n\n```bash\ncd bliss\npoetry install\npoetry shell\n```\n\n6. Verify that bliss is installed correctly by running the tests both on your CPU (default) and on your GPU:\n\n```bash\npytest\npytest --gpu\n```\n\n7. Finally, if you are planning to contribute code to this repository, consider installing our pre-commit hooks so that your code commits will be checked locally for compliance with our coding conventions:\n\n```bash\npre-commit install\n```\n\n# Latest updates\n## Galaxies\n - BLISS now includes a galaxy model based on a VAE that was trained on Galsim galaxies.\n - BLISS now includes an algorithm for detecting, measuring, and deblending galaxies.\n\n## Stars\n - BLISS already includes the StarNet functionality from its predecessor repo: [DeblendingStarFields](https://github.com/Runjing-Liu120/DeblendingStarfields).\n\n\n# References\n\nMallory Wang, Ismael Mendoza, Cheng Wang, Camille Avestruz, and Jeffrey Regier. *Statistical Inference for Coadded Astronomical Images.* Machine Learning and the Physical Sciences workshop, NeurIPS 2022. [arXiv:2211.09300](https://arxiv.org/abs/2211.09300)\n\nDerek Hansen, Ismael Mendoza, Runjing Liu, Ziteng Pang, Zhe Zhao, Camille Avestruz, and Jeffrey Regier. *Scalable Bayesian Inference for Detection and Deblending in Astronomical Images*. ICML Workshop on Machine Learning for Astrophysics, 2022. [arXiv:2207.05642](https://arxiv.org/abs/2207.05642)\n\nRunjing Liu, Jon D. McAuliffe, Jeffrey Regier, and The LSST Dark Energy Science Collaboration. *Variational Inference for Deblending Crowded Starfields*, 2021. [arXiv:2102.02409](https://arxiv.org/abs/2102.02409)\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Bayesian Light Source Separator",
"version": "0.3",
"project_urls": {
"Documentation": "https://prob-ml.github.io/bliss/",
"Homepage": "https://github.com/prob-ml/bliss",
"Repository": "https://github.com/prob-ml/bliss"
},
"split_keywords": [
"cosmology",
"blending",
"weak lensing",
"bayesian",
"ml",
"pytorch"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "166907c8dec44a50115529d5db1195fdcd0163da0eef4d41262bf1f34e47cd4a",
"md5": "206187d99d69aec5e604ffbb0fa8a3b9",
"sha256": "ef5cdcd13fb0d27477dd046a4542bbed445a3afa24c407b58ab8a2713b7eb512"
},
"downloads": -1,
"filename": "bliss_toolkit-0.3-py3-none-any.whl",
"has_sig": false,
"md5_digest": "206187d99d69aec5e604ffbb0fa8a3b9",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.10,<4.0",
"size": 77559,
"upload_time": "2023-08-10T05:54:47",
"upload_time_iso_8601": "2023-08-10T05:54:47.998997Z",
"url": "https://files.pythonhosted.org/packages/16/69/07c8dec44a50115529d5db1195fdcd0163da0eef4d41262bf1f34e47cd4a/bliss_toolkit-0.3-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "1742d32d35deff38cb04d8526106f444346a3e4ffe74dbbe553313c4ab24a165",
"md5": "4dccc7511905de3e0bf9c45b17c19341",
"sha256": "1b548be57a7fb48499aa0c922380cdf42f6b3f4d6ad69a582429fc0471f0b22f"
},
"downloads": -1,
"filename": "bliss_toolkit-0.3.tar.gz",
"has_sig": false,
"md5_digest": "4dccc7511905de3e0bf9c45b17c19341",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.10,<4.0",
"size": 66707,
"upload_time": "2023-08-10T05:54:49",
"upload_time_iso_8601": "2023-08-10T05:54:49.929196Z",
"url": "https://files.pythonhosted.org/packages/17/42/d32d35deff38cb04d8526106f444346a3e4ffe74dbbe553313c4ab24a165/bliss_toolkit-0.3.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-08-10 05:54:49",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "prob-ml",
"github_project": "bliss",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "bliss-toolkit"
}
JSON
